Journal Issue
Share
Article

Colombia: Selected Issues Paper

Author(s):
International Monetary Fund. Western Hemisphere Dept.
Published Date:
June 2015
Share
  • ShareShare
Show Summary Details

Unemployment-Growth Trends in Colombia and Selected Emerging Markets1

1. The relationship between unemployment and economic growth differs vastly across emerging markets (EMs). This relationship is strongest in EM Europe, where the global financial crisis (GFC) led to a sizable increase in the unemployment rate. And while unemployment has gradually decreased in line with the recovery, the level of unemployment remains above pre-crisis levels. The picture is different in Latin America, where unemployment continued on a strong downward trend, after a small and short-lived increase in unemployment during the GFC, despite lower post-crisis growth. The continued decline possibly suggests that structural rather than cyclical factors drive recent labor market improvements. In Asia and the Middle East, the relationship between economic growth and labor markets remains extremely week, although data availability in these countries is very limited and often only covers major cities.

Emerging Markets, Europe

Source: IFS, Datastream and Haver.

Emerging Markets, Latin America

Source: IFS, Datastream and Haver.

Emerging Markets, Asia

Source: IFS, Datastream and Haver.

Emerging Markets, Middle East & Central Asia

Source: IFS, Datastream and Haver.

2. Okun’s coefficients are estimated for selected EMs to measure differences in the unemployment-growth relationship. Okun’s coefficients—first estimated by Okun in 1962 for the U.S.—capture the sensitivity of the unemployment rate to cyclical fluctuations in output. While the coefficients generally range between 0.3 and 0.4 for advanced economies, the EM average is significantly lower. The 2010 October WEO, for instance, finds that these coefficients for EMs average at about 0.10. We extend and refine this work by estimating dynamic country-specific Okun’s coefficients for a larger set of EMs using seasonally adjusted quarterly data for 2000Q1 to 2014Q4 (see text chart, non-significant coefficients are reported as zeros).2 In line with the data trends described above, the Okun’s coefficients are largest for European EMs, followed by Latin American EMs. No significance is found for Asian EMs, and in the Middle East, only Egypt displays sensitivity of unemployment to cyclical movements in the expected direction. In the case of Colombia, the unemployment rate decreases by 0.16 percentage points when GDP increases by one percent.

Unemployment and Growth (Okun’s coefficients)

3. The large heterogeneity in these coefficients across countries and regions can be attributed to country specific labor market and economic structures. Comparing Okun’s coefficients across countries and along various institutional and policy variables suggests that countries with greater degrees of informality and labor market rigidities (after excluding Asia and Middle East) tend to have to smaller Okun’s coefficients. Meanwhile, education, a well developed services sector, and trade openness tend to be correlated with higher sensitivity of labor markets.

Okun Coefficient and Informal Sector Employment 1/

1/ Okun coefficient that is not statistically significant is replaced with value of zero.

Okun Coefficient and Secondary Education of Labor Force1/

1/ Okun coefficient that is not statistically significant is replace with value of zero.

4. While Colombia’s Okun’s coefficient appears broadly in line with country characteristics, the sensitivity of labor markets to the cycle has decreased over time. The Okun’s coefficient is somewhat lower than for its regional peers, but the higher degree of informality and the relatively less open economy could help explain this small difference. However, there is some evidence that the sensitivity of labor markets to output fluctuations may be falling (REO 2014). This is puzzling as informality has declined significantly in the past years while education levels and the service sector have strengthened. Both developments tend to be correlated with an increase of unemployment sensitivity to cyclical fluctuations. On the other hand, some measures of labor market flexibility, notably wage determination, have slightly worsened, which could have made markets more rigid and less able to adjust to the economic cycle.

Informality and Formality Rates

(Percent, 13 metropolitan areas, 3MMA)

Source: Dane.

Labor Market Flexibility

A. Conclusion

5. The sensitivity of unemployment displays strong heterogeneity across countries. The degree of informality, labor market rigidities as well as openness and education play an important role in explaining those differences. While Colombia’s Okun’s coefficient is relatively small, it is broadly in line with cross-country experience. Empirical evidence suggests that the sensitivity of unemployment to cyclical conditions has decreased recently. While there is no normatively ‘optimal’ Okun’s coefficient, a country with a favorable medium term growth outlook such as Colombia would rather benefit from a relatively strong sensitivity of unemployment to output. Policies which help translate economic momentum into improvements in labor markets would therefore be helpful, including further reducing informality, strengthening education, and removing labor market rigidities.

References

    ArangoL. and F.Hamann (Editors)2013El Mercado de trabajo en Colombia: hechos, tendencias e instituciones” (Bogotá: Banco de la República).

    • Search Google Scholar
    • Export Citation

    International Monetary Fund2010World Economic Outlook: Rebalancing GrowthApril (Washington).

    International Monetary Fund2014Regional Economic Outlook Update: Western Hemisphere DepartmentOctober (Washington).

    OkunA.1962Potential Output: Its Measurement and Significance,Proceedings of the Business and Economic Statistics Section of the American Statistical Society.

    • Search Google Scholar
    • Export Citation
1Prepared by Christina Kolerus (SPR). We would like to thank Luis Cubeddu (SPR) for the helpful comments.
2The optimal lag structure for both GDP and unemployment was determined using the Akaike’s Information Criterion.

Other Resources Citing This Publication